Pseudomonas fluorescens - ACS Publications - American Chemical ...

Article pubs.acs.org/EF

Transesterification of Palm Oil Catalyzed by Pseudomonas fluorescens Lipase in a Packed-Bed Reactor Gisanara Dors,† Larissa Freitas,‡ Adriano A. Mendes,§ Agenor Furigo, Jr.,† and Heizir F. de Castro*,‡ †

Federal University of Santa Catarina, Post Office Box 476, 88010-970 Florianópolis, Santa Catarina, Brazil Engineering School of Lorena, University of São Paulo, Post Office Box 116, 12602-810 Lorena, São Paulo, Brazil § Federal University of São João del Rei, 35701-970 Sete Lagoas, Minas Gerais, Brazil ‡

ABSTRACT: Transesterification of palm oil with ethanol catalyzed by Pseudomonas fluorescens lipase immobilized on epoxy− polysiloxane−polyvinyl alcohol composite (epoxy−SiO2−PVA) was performed in a continuous packed-bed reactor (PBR). Two strategies were used for improving the miscibility of the substrates: the addition of the organic solvent tert-butanol and the surfactant Triton X-100. Results were compared to those obtained in a solventless reactor, which displayed a biphasic system that passed through the reactor. Using this system, the ethyl ester yield of 61.6 ± 1.2% was obtained at steady state. Both Triton X100 and tert-butanol systems were found to be suitable to promote the miscibility of the starting materials; however, the use of Triton X-100 reduced the yield to levels lower than 20%, because of the enzyme desorption from the support surface, as confirmed by scanning electron microscopy analysis. The best performance was found for the reactor running in the presence of tert-butanol, which resulted in a stable operating system and an average yield of 87.6 ± 2.5%. This strategy also gave high biocatalyst operational stability, revealing a half-life of 48 days and an inactivation constant of 0.6 × 10−3 h−1.

■

INTRODUCTION Conventionally, the production of industrial biodiesel is based on the chemical transesterification of vegetable oils with methanol, using homogeneous catalysts to promote the cleavage of the triglyceride molecule and generate a mixture of fatty acid alkyl esters (biodiesel) and glycerol.1 In this process, high molar conversion can be obtained at short reaction times; however, its energy consumption is high and requires significant downstream processing steps, such as washing, separation, and purification.2,3 Alternatively, transesterification of triglycerides to produce biodiesel can be obtained enzymatically using lipase (Scheme 1). Despite the

activity and yield in non-aqueous media, and temperature and alcohol resistance, and establishing optimal conditions, have all been solved at a laboratory-research level.4,6,7 However, a suitable process technology has yet to be established. Given our background in developing strategies to improve the transesterification yields and to make the reaction time as short as possible using different feedstocks under batch process,8−10 it would be logical to continue this research by testing the process running on a continuous basis. For this, a packed-bed reactor (PBR) configuration was selected on the basis of its suitability to perform typical lipase-catalyzed reactions.11−14 PBRs are kinetically more favorable than continuous stirred tank reactors (CSTRs) because the disadvantage of the high mechanical stress caused by the agitation can be avoided.13 However, this reactor configuration was found to have some constraints in attaining satisfactory performance when running on vegetable oil with high levels of saturated fatty acids, such as palm and babassu oils.8,9 This was related to the chemical composition available to the biocatalyst caused by the immiscibility of the starting materials, obstruction of the packed bed resulting in channeling, and lower yields compared to batchwise processes. In addition, immobilized enzymes are densely packed, thereby leading to considerable difficulties in removing the highly viscous and hydrophilic glycerol byproduct. Glycerol tends to adsorb on enzyme immobilization carriers and forms a hydrophilic layer, which makes lipases inaccessible to hydrophobic substrates.14 To overcome these limitations, two strategies were tested to make the reaction medium miscible: the addition of the surfactant (Triton X-100) and the solvent (tert-butanol). Here,

Scheme 1. Transesterification Reaction of Triacylglycerols with Short-Chain Alcohols as Acyl Acceptors

challenges involving the use of enzymes on an industrial scale, this approach appears to be a promising alternative to chemically catalyzed biodiesel production, because the enzymatic process consumes less energy, produces less waste and byproducts, and requires mild conditions.4 Indeed, the enzymatic synthesis of biodiesel has become an attractive research topic, and publications are numerous.5−7 In principle, the biochemical problems, such as finding a lipase with the desired characteristics, including the ability to use all mono-, di-, and triglycerides, low product inhibition, high © 2012 American Chemical Society

Received: May 29, 2012 Revised: August 1, 2012 Published: August 1, 2012 5977

dx.doi.org/10.1021/ef300931y | Energy Fuels 2012, 26, 5977−5982

Energy & Fuels

Article

we report the results, taking the enzymatic transesterification of palm oil with ethanol catalyzed by Pseudomonas fluorescens lipase immobilized on epoxy−polysiloxane−polyvinyl alcohol composite (epoxy−SiO2−PVA) as the reaction model. Data were compared to those attained in the reactor running in a solvent-free system under the same operational conditions. The enzymatic transesterification using ethanol instead of methanol has been suggested as a cleaner and more sustainable alternative for biodiesel production.15−17 Furthermore, ethanol is a larger and heavier alcohol than methanol, which means a mass yield gain in the enzymatic synthesis of fatty acid ethyl esters (FAEEs), resulting in a higher biodiesel per unit of oil value.18 In some countries, such as Brazil, ethanol is sold at lower prices than methanol, which means that the alcohol component is always significantly cheaper than the oil component. Thus, the extra volume gain when using ethanol instead of methanol could become a major sales argument, particularly for the Brazilian market.15,19

■

Figure 1. Schematic diagram of the experimental apparatus: (1) thermostatic bath, (2) magnetic stirrer, (3) substrate reservoir, (4) reflux condenser, (5) peristaltic pump, (6) PBR, and (7) product outlet.

EXPERIMENTAL SECTION

Materials. Refined and bleached palm oil was a kind gift from Agropalma (Belem, Pará, Brazil), having the following composition in fatty acids: 0.1 wt % lauric acid, 1.2 wt % myristic acid, 46.8 wt % palmitic acid, 3.8 wt % stearic acid, 37.6 wt % oleic acid, and 10.5 wt % linoleic acid, with 864.2 g/mol average molecular weight. Other characteristics of the palm oil sample include: acid value, 0.33 mg of KOH/g; peroxide value, 2.05 meq/kg; iodine value, 98 g of I2/g; and saponification value, 198 mg of KOH/g. Ethanol (minimum of 99%) and tert-butanol were from Cromoline (São Paulo, Brazil). Triton X100 was acquired from Sigma-Aldrich (St. Louis, MO). Tetraethoxysilane (TEOS) and polyvinyl alcohol (PVA, MW of 88 000) were acquired from Aldrich Chemical Co. (Milwaukee, WI). Hydrochloric acid (minimum of 36%) and polyethylene glycol (PEG, 1500 g/mol) were supplied by Synth (São Paulo, Brazil). Commercial olive oil (low acidity, purchased in a local market) was used to determine the hydrolytic activity. All of the other reagents were of analytical grade. Immobilized Biocatalyst. Immobilized lipase from P. fluorescens (Lipase AK, 20715TD from Amano Pharmaceuticals, Nagoya, Japan) supported on epoxy−SiO2−PVA was prepared following previously established methodology.20,21 The properties of the support were as follows: diameter, 0.175 mm; average pore diameter, 22.91 Å; surface area, 461 m2/g; and porous volume, 0.275 cm3/g.21 To perform this work, four batches of immobilized derivative were prepared, and the average measured hydrolytic activity was 2250 ± 194 units/g according to the previously described methodology.22 A total of 1 unit of enzyme activity was defined as the amount of enzyme that liberates 1 μmol of free fatty acid per minute under the assay conditions (37 °C, pH 7.0, and 150 rpm). The biochemical, kinetic properties, thermal stability, and operational stability of this immobilized lipase preparation are described elsewhere.18 Continuous Runs. The transesterification reactions were carried out in the PBR jacketed glass column (internal diameter, 160 mm; height, 55 mm; and total volume, 11 mL) with a water jacket connected to a circulating water bath to maintain the temperature at 50 °C. A schematic diagram of the PBR is shown in Figure 1. The continuous run was started by loading the reactor with the biocatalyst, and the substrate was continuously pumped (peristaltic pump Perista Pump SJ-1211, Atto Bioscience and Biotechnology, Tokyo, Japan) from a reservoir, through marprene tubing, to the bottom end of the bioreactor at different flow rates. The reflux condenser system was connected to the feeding vessel to avoid ethanol losses. Heating tapes containing a thermostatted electrical resistance (25 W) were used to avoid heat loss in the inlet and outlet tubing. For each run, an amount of 6.7 g of biocatalyst was used, which corresponds to a bulk volume of 9.97 cm3. Immobilized derivative density was determined as 1.865 g/ mL. The space time for runs 1−3 was 4.25 h, and for run 4, the space time ranged from 3.7 to 5.5 h.

Runs were performed using a substrate composed of ethanol/palm oil at a molar ratio of 9:1 and designed as run 1, solvent-free system; run 2, substrate containing 10% Triton X-100 (in relation to the total volume); and runs 3 and 4, substrate in the presence of tert-butanol (30% in relation to the oil mass). The runs 1 and 3 had a duration of 240 h, while run 2 lasted 168 h. Samples were collected each 24 h during the operation of the PBR, diluted in n-hexane, and stored at −2 °C to carry out the chromatograph analyses. Each data point was repeated 3 times, and the measurement error was lower than 5%. The biocatalyst stability was assessed by measuring the hydrolytic activity of the immobilized derivatives at the end of each continuous run (runs 1−4), taking the original activity as 100%. The recovered immobilized lipase was then washed with tert-butanol to remove any substrate or product eventually retained in the matrix. Hydrolytic activity was determined by the olive oil emulsion method according to the modification proposed by Soares et al.22 The inactivation constant (kd) and half-life (t1/2) for the immobilized lipase were calculated according to the eqs 1 and 2, as follows:

ln A t = ln A 0 − kdt t1/2 =

ln 2 kd

(1) (2)

where A0 is the initial activity of the immobilized lipase and At is the final activity after each run. Monitoring Ethyl Esters. The ethyl esters formed in the transesterification reaction were analyzed in flame ionization detector (FID) gas chromatography (Varian CG 3800, Inc., Palo Alto, CA) using a 5% DEGS CHR-WHP, 80/100-mesh, 6 ft, 2.0 mm inner diameter, and 1/8 in. outer diameter column (Restek, Frankel Commerce of Analytic Instruments, Ltd., São Paulo, Brazil), following previous established conditions.23 Nitrogen was used as the carrier gas with a flow rate of 25 mL/min. The detector and injector temperatures were 190 °C. The column temperature was first set to 90 °C for 3 min and then programmed at 25 °C/min to 120 °C for 10 min and 170 °C for 15 min. Data were collected using Galaxie Chromatography Data System software, version 1.9. Theoretical ester concentrations were calculated by taking into account the palm oil fatty acid composition and its initial weight mass in the reaction medium,9,23 and the transesterification yield (%) was defined as the ratio between the produced and theoretical ester concentrations.9 Scanning Electron Microscopy (SEM). The morphology and the surface of the immobilized derivative samples before and after trials 5978


Energy & Fuels

Article

suggested by Valerio et al.24 The authors verified that the use of emulsifiers, such as Triton X-100 and Tween 65, at high concentrations (up to 16%) gave the highest yield in monoand diacylglycerols in the enzymatic glycerolysis of olive oil mediated by Novozym 435 in a batch system. At 24 h, the measured concentration of ethyl esters was higher than runs performed in the absence and presence of an organic solvent (runs 1 and 3). After this, the ethyl ester yield markedly decreased to a level lower than 20%. This drop may be associated with the lipase desorption from the support SiO2− PVA, because of the action of the surfactant. This hypothesis was checked by comparing the morphological structure of both original and recovered immobilized derivatives through SEM, as shown in panels a and b of Figure 3.

were examined by SEM. Samples were sputter-coated with gold and scanned at 10−15 kV by SEM Zeiss LEO (model 1450VP).

■

RESULTS AND DISCUSSION Evaluation of the Tested Strategies. Microbial lipase from P. fluorescens immobilized on hybrid matrix epoxy−SiO2− PVA showed high activity in the transesterification of palm oil with ethanol, attaining high yields in batch runs.9 After successful enzyme immobilization,20 the development of an appropriate reactor is a critical factor in enzymatic biodiesel production on an industrial scale. The enzymatic transesterification of vegetable oils can be carried out substantially faster and more economically in continuous reactors. In the PBR system, the substrate is directly passed on the catalyst layer, thereby increasing the contact surface of the substrate and catalyst. The transesterification of palm oil was performed in a solvent-free system as the control (run 1), in the presence of surfactant Triton X-100 (run 2), and in the presence of tertbutanol (run 3). The addition of Triton X-100 has been reported to be a successful strategy to increase the lipase activity in the glycerolysis of olive oil using Novozym 435 as a catalyst in a batch system.24 Taking into account that P. fluorescens lipase is not active toward tertiary alcohols,25 tertbutanol has been used to increase the miscibility of substrates and, consequently, enhance the transesterification reaction rates catalyzed by immobilized lipases under batch and continuous reactors.26,27 Results attained in terms of ethyl ester yield as a function of time are displayed in Figure 2.

Figure 3. Scanning electron micrographs: (a) original biocatalyst before its use, with magnification at 200×, beam energy at 10 kV, and scale at 100 μm, and (b) biocatalyst after its use (run 2, substrate with the addition of Triton X-100 at 10 wt %), with magnification at 180×, beam energy at 15 kV, and scale at 100 μm.

Figure 2. Ethyl ester yield in the continuous transesterification of palm oil catalyzed by P. fluorescens lipase immobilized on SiO2−PVA carried out in the PBR. Run 1, solventless substrate (□); run 2, substrate with the addition of Triton X-100 (△); and run 3, substrate containing tertbutanol (●).

Before its use (Figure 3a), the immobilized derivative showed a typical flat support surface tightly compacted by the enzyme. After its use, the enzyme layer was completely desorbed from the support (Figure 3b), which probably helped it wash out from the reactor. This behavior agrees with the attained results (Figure 2), showing a gradual decrease in the ethyl ester concentrations until no ester formation was detected at 168 h. Similar data were previously described for Candida antarctica (CALB) immobilized on different supports.28 According to this investigation, immobilized CALB derivatives were fully desorbed from hydrophobic supports after incubating the biocatalysts with Triton X-100. The suitable concentration of the surfactant to fully desorb the enzyme was found to be dependent upon the support type, varying from 1% for butylagarose to 4% for octadecyl-Sepabeads. The effect of the lipase

For the control assay (run 1), the reactants within the bed form a two-phase liquid system, despite stirring in the substrate reservoir. The steady state was reached at 72 h, and the average ethyl ester yield was 61.6 ± 1.2%. The addition of either Triton X-100 or tert-butanol helped to reduce by 50% the time needed to achieve the steady state; however, different performances were observed. Triton X-100, a non-ionic surfactant added at a concentration of 10%, was found to be enough to turn the oil/ethanol system miscible. This proportion was lower than the one 5979


Energy & Fuels

Article

presence of tert-butanol (run 3), a slight increase on the biocatalyst half-life value (t1/2 = 277 h) was observed. Therefore, among the strategies evaluated in the present study, the continuous production of ethyl esters in the presence of tert-butanol as the solvent (run 3) was found to have the best performance and also helped to increase the biocatalyst half-life. Influence of the Space Time in the Transesterification of Palm Oil in the tert-Butanol System Performed in the PBR. To confirm the performance of the strategy selected, a long-term operation was carried out (run 4) feeding the reactor with a substrate containing ethanol, palm oil, and tert-butanol. The substrate flow rate varied from 1.2 to 1.8 mL/h, corresponding to space times in the range of 5.5−3.7 h, respectively. Figure 4 displays the dynamic behavior of the ethyl ester concentration, a product of the reaction, along with the duration of the experiment (528 h).

desorption from the support was also reported for lipase from Candida rugosa immobilized onto zirconium phosphate when used in sequential batch runs [hydrolysis of p-nitrophenilpalmitate (p-NPP) emulsified with Triton X-100].29 After 12 cycles, the hydrolytic activity of the biocatalyst decreased 80% in relation to the initial activity (t1/2 = 0.93 h), which again is in agreement with the results attained in the present study. On the other hand, the addition of tert-butanol as a solvent was found to enhance the ethyl ester production in the PBR. The ester concentration attained under steady-state conditions corresponded to a transesterification yield of 81.6 ± 5.4%. This enhanced performance cannot only be attributed to the miscibility of the starting materials but also to specific properties of tert-butanol, whose moderate polarity can dissolve both formed products (ethyl esters) and byproduct (glycerol). It also appears that the formed ethyl ester can act as a mutual solvent for the reactants, thus helping to generate a single-phase system. Simultaneously, the glycerol formed (byproduct) has a positive effect on the chemical equilibrium between the formation of ethyl esters, as already observed by Chen et al.30 Therefore, the high ethyl ester formation can be credited not only to the ability of tert-butanol to dissolve glycerol but also from its enhancement of oil−ethanol miscibility, which overcomes the mass-transfer restraints as well. This enables a stable reactor operation in relation to the ester concentration under steady-state conditions for up to 10 days, as well as minimizing the effect of inhibition on the biocatalyst. The use of tert-butanol as the solvent in the synthesis of several components in lab scale has been widely reported in the literature.26,27,31,32 At the industrial level, this strategy has also been used to produce 1 ton/h of biodiesel in a continuous transesterification of waste/used vegetable oil with methanol mediated by Novozym 435.33 Further information on the reactor behavior was given in terms of the biocatalyst half-life by measuring the enzymatic activity at the beginning and end of each run. Table 1 shows the values for enzymatic activity, the inactivation constant (kd), and the biocatalyst half-life (t1/2).

Figure 4. Ethyl ester concentrations from transesterification of palm oil performed in the PBR using immobilized P. fluorescens. The reactions were carried out at a fixed molar ratio of ethanol/oil (9:1), with 30% (v/v) tert-butanol at 50 °C, feed flow rate from 1.2 to 1.8 mL/h, and space time in the range of 5.5−3.7 h.

Table 1. Half-Life (t1/2) and Inactivation Constant (kd) for P. fluorescens Lipase Immobilized on Epoxy−SiO2−PVA Estimated in the Continuous Transesterification of Palm Oil Carried out in the PBR under Different Strategiesa run

time (h)

1 2 3 4

240 168 240 528

initial activity (units/g) 2125 2180 2560 2322

± ± ± ±

83 20 78 36

final activity (units/g) 963 475 1398 1680

± ± ± ±

23 18 64 62

operational stability, t1/2 (h)

inactivation constant, kd (×10−3, h−1)

210 76 277 1155

3.3 9.1 2.5 0.6

From the analysis of Figure 4, it can be verified that, for the lowest flow rate (1.2 mL/h), the steady state was reached in 48 h, lasting approximately another 96 h and reaching ethyl ester concentrations in the order of 56.8 ± 1.8 wt %. A similar ethyl ester concentration (54.3 ± 0.7 wt %) was obtained at a flow rate of 1.5 mL/h, corresponding to a space time of 4.6 h. Increasing the substrate flow rate to 1.8 mL/h decreases the ethyl ester concentration to 36.9 ± 3.2 wt %. To investigate the relationship of space time with the resulting biodiesel conversion, the average values for productivity and concentration of ethyl esters are shown in Figure 5. When the flow rate increased from 1.2 to 1.5 mL/h, an increase in productivity from 93 to 119 mgethyl esters g−1 h−1 was observed without a significant change on the ester concentration. A further increase in the flow rate (1.8 mL/h) caused a fall in both the productivity (98 mgethyl esters g−1 h−1) and the ethyl ester concentration (36.9 ± 3.2 wt %). As shown in Figure 5, the best performance for the PBR running on palm oil was found at a space time of 4.6 h (flow rate of 1.5 mL/h), attaining high productivity (119 mgethyl esters g−1 h−1) and yield on ethyl esters (87.6 ± 2.5%). In addition, the residual hydrolytic activity of the immobilized lipase was estimated at 1680 units/g, which corresponded to a 28% loss of

a

Hydrolytic activity was determined by the olive oil emulsion method according to the modification proposed by Soares et al.22 A total of 1 unit of enzyme activity was defined as the amount of enzyme that liberates 1 μmol of free fatty acid per minute under the assay conditions (37 °C, pH 7.0, and 150 rpm).

The addition of Triton X-100 drastically reduced the catalytic activity of the immobilized lipase because of the enzyme leakage from the support (Figure 3). Under these conditions, the immobilized lipase revealed a half-life of 76 h, which was almost 3-fold lower than the value determined for the control assay in run 1 (t1/2 = 210 h). For the reaction carried out in the 5980


Energy & Fuels

Article

and 70% of the biocatalyst activity was retained even after continuous operation for almost 48 days.

■

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS The authors gratefully acknowledge the financial support of ́ Conselho Nacional de Desenvolvimento Cientifico e Tecnológico (CNPq) and Coordenaçaõ de Aperfeiçoamento de Pessoal ́ Superior (CAPES). de Nivel

■

Figure 5. Productivity and ethyl ester concentration for the transesterification of palm oil performed in the PBR using immobilized P. fluorescens. The reactions were carried out at a fixed molar ratio of ethanol/oil (9:1), with 30% (v/v) tert-butanol at 50 °C, feed flow rate from 1.2 to 1.8 mL/h, and space time in the range of 5.5−3.7 h.

REFERENCES

(1) Quintella, C. M.; Teixeira, L. S. G.; Korn, M. G. A.; Neto, P. R. C.; Torres, E. A.; Castro, M. P.; Jesus, C. A. C. Quim. Nova 2009, 32, 793−808. (2) Knothe, G.; Van Gerpen, J.; Krahl, J. The Biodiesel Handbook; American Oil Chemists’ Society (AOCS) Press: Urbana, IL, 2005. (3) Pinto, A. C.; Guarieiro, L. L. N.; Rezende, M. J. C.; Ribeiro, N. M.; Torres, E. A.; Lopes, W. A.; Pereira, P. A. D.; Andrade, J. B. J. Braz. Chem. Soc. 2005, 16, 1313−1330. (4) Ghaly, A. E.; Dave, D.; Brooks, M. S.; Budge, S. Am. J. Biochem. Biotechnol. 2010, 6, 54−76. (5) Nie, K. L.; Xie, F.; Wang, F.; Tan, T. W. J. Mol. Catal. B: Enzym. 2006, 43, 142−147. (6) Nielsen, P. M.; Brask, J.; Fjerbaek, L. Eur. J. Lipid Sci. Technol. 2008, 110, 692−700. (7) Parawira, W. Crit. Rev. Biotechnol. 2009, 29, 82−93. (8) Paula, A. V.; Urioste, D.; Santos, J. C.; de Castro, H. F. J. Chem. Technol. Biotechnol. 2007, 82, 281−288. (9) Moreira, A. B. R.; Perez, V. H.; Zanin, G. M.; de Castro, H. F. Energy Fuels 2007, 21, 3689−3694. (10) Freitas, L.; Da Rós, P. C. M.; Santos, J. C.; de Castro, H. F. Process Biochem. 2009, 44, 1068−1074. (11) Fernandes, P. Enzyme Res. 2010, DOI: 10.4061/2010/862537. (12) Santori, G.; Di Nicola, G.; Moglie, M.; Polonara, F. Appl. Energy 2012, 92, 109−132. (13) Freitas, L.; Santos, J. C.; Zanin, G. M.; de Castro, H. F. Appl. Biochem. Biotechnol. 2010, 161, 372−381. (14) Séverac, E.; Galy, O.; Turon, F.; Pantel, C. A. Enzyme Microb. Technol. 2011, 48, 61−70. (15) Stamenkovic, O. S.; Velickovic, A. V.; Veljkovic, B. V. Fuel 2011, 90, 3141−3155. (16) Souza, M. S.; Aguieiras, E. C. G.; Silva, M. A. P.; Langone, M. A. P. Appl. Biochem. Biotechnol. 2009, 154, 74−88. (17) Véras, I. C.; Silva, F. A. L.; Ferrão-Gonzales, A. D.; Moreau, V. H. Bioresour. Technol. 2011, 9653−9658. (18) Vieitez, I.; Silva, C.; Alckmin, I.; Borges, G. R.; Corazza, F. C.; Oliveira, J. V.; Grompone, M. A.; Jachmanián, I. Renewable Energy 2010, 35, 1976−1981. (19) Brunschwing, C.; Moussavou, W.; Blin, J. Prog. Energy Combust. Sci. 2012, 38, 283−301. (20) Santos, J. C.; Paula, A. V.; Nunes, G. F. M.; de Castro, H. F. J. Mol. Catal. B: Enzym. 2008, 52−53, 49−57. (21) Santos, J. C.; Paula, A. V.; Rocha, C. G. F.; Nunes, G. F. M.; de Castro, H. F. J. Non-Cryst. Solids 2008, 354, 4823−4826. (22) Soares, C. M. F.; de Castro, H. F.; Moraes, F. F.; Zanin, G. M. Appl. Biochem. Biotechnol. 1999, 77/79, 745−757. (23) Urioste, D.; Castro, M. B. A.; Biaggio, F. C.; de Castro, H. F. Quim. Nova 2008, 31, 407−412. (24) Valério, A.; Rovani, S.; Treichel, H.; Oliveira, D.; Oliveira, J. V. Bioprocess Biosyst. Eng. 2010, 33, 805−812. (25) Henke, E.; Pleiss, J.; Bornscheuer, U. T. Angew. Chem., Int. Ed. 2002, 4, 3211−3213.

activity compared to its initial activity (2322 units/g), revealing a half-life (t1/2) of 48 days (Table 1). In agreement with the data described in the literature, the use of organic solvents in the enzymatic synthesis of biodiesel increases the reaction rate, starting from a mutual improvement in the solubility of the hydrophobic triglycerides and alcohols,34 as well as improving the operational stability of the lipase preparations.35 Using similar reaction conditions, the performance in the PBR system proved to have a better profile than in a batch stirred tank reactor.9 Under batch conditions, almost complete conversion of the oil to ethyl esters (97.9%) was achieved in 24 h (productivity of 23.9 mgethyl esters g−1 h−1).9 In contrast, the continuous run allows for 87.6 ± 2.5% of oil conversion to be attained in 4.5 h. Hence, the continuous trials (119 mgethyl esters g−1 h−1) were about 5 times higher than batch runs. The necessity of solvent recovery can be a drawback to this strategy. However, several positive aspects can balance such limitations:26 the tert-butanol concentration for optimum conversion is not high, and consequently, the energy expense required for its recovery can be acceptable; solvent recovery is a common practice in the chemical-catalyzed production of biodiesel, and it is necessary in all cases to remove the excess alcohol; and the low boiling point of tert-butanol makes for an easy separation of the solvent together with the ethanol.

■

CONCLUSION The objective of this work was to develop an efficient system for continuous enzymatic synthesis of biodiesel from palm oil and ethanol in the PBR using P. fluorescens lipase immobilized on epoxy−SiO2−PVA. Our challenge was to overcome some limitations for PBR operation because of the immiscibility of the reactants (ethanol and oil), which caused obstruction of the bed, resulting in channeling and lower yields compared to processes running on batch criteria. Of the two strategies tested, the use of Triton X-100 was found to be unsuitable because of the strong enzyme leakage from the support surface, as confirmed by SEM analysis. The best approach was the use of tert-butanol as a solvent, which resulted in a miscible system and high yield values. Under these conditions, the steady-state yield of 87.6 ± 2.5% was attained at the flow rate of 1.5 mL/h 5981


Energy & Fuels

Article

(26) Royon, D.; Daz, M.; Ellenrieder, G.; Locatelli, S. Bioresour. Technol. 2007, 98, 648−653. (27) Fu, B.; Vasudevan, P. T. Energy Fuels 2010, 24, 4646−4651. (28) Cabrera, Z.; Fernandez-Lorente, G.; Fernandez-Lafuente, R.; Palomo, J. M.; Guisan, J. M. J. Mol. Catal. B: Enzym. 2009, 57, 171− 176. (29) Mendes, A. A.; Barbosa, B. C. M.; Silva, M. L. C. P.; de Castro, H. F. Biocatal. Biotransform. 2007, 25, 393−400. (30) Chen, Y. H.; Huang, Y. H.; Lin, R. H.; Shang, N. C. Bioresour. Technol. 2010, 101, 668−673. (31) Chang, C.; Chen, J. H.; Chang, C. M. J.; Wu, T. T.; Shieh, C. J. New Biotechnol. 2009, 26, 187−192. (32) Raita, M.; Champreda, V.; Laosiripojana, N. Process Biochem. 2010, 45, 829−834. (33) Al-Zuhair, S.; Almenhali, A.; Hamad, I.; Alshehhi, M.; Alsuwaidi, N.; Mohamed, S. Renewable Energy 2011, 36, 2605−2614. (34) Soumanou, M. M.; Bornscheuer, U. T. Enzyme Microb. Technol. 2003, 33, 97−103. (35) Antczak, M. S.; Kubiak, A.; Antczak, T.; Bielecki, S. Renewable Energy 2009, 34, 1185−1194.

5982
